Defeat device testing in the EU: So far, not so good

Posted Thursday, 28 April 2016, 14:02

Conclusively determining whether or not a vehicle has a defeat device to circumvent emissions tests is difficult. Really difficult. That is because a defeat device, as defined in EU regulation, is:

any element of design which senses temperature, vehicle speed, engine speed (RPM), transmission gear, manifold vacuum or any other parameter for the purpose of activating, modulating, delaying or deactivating the operation of any part of the emission control system, that reduces the effectiveness of the emission control system under conditions which may reasonably be expected to be encountered in normal vehicle operation and use [emphasis added]

This means that there are two key elements to a defeat device: (1) sensing the vehicle is in normal operation, as opposed to under laboratory test conditions; and (2) reducing the effectiveness of the emissions control system accordingly. In order to conclusively prove the existence of a defeat device, it is not enough to show that the vehicle produces high real-world emissions. You must also show how the vehicle sensed it was in normal operation and how the emissions control was changed as a result.

The UK, France, and Germany have all just reported results of defeat device testing programs they commenced in the wake of the VW revelations last September. The testing protocols varied, but all consisted of comparing variations of real-world and laboratory testing. The French protocol involved three different tests, the UK protocol seven, and the German protocol eight (see table). These testing protocols should all be considered screening tests. In other words, the results can be used to identify likely offenders that would be good candidates for further investigation. But these testing protocols are not sufficient on their own to conclusively prove or disprove the presence of a defeat device.

Summary of defeat device testing protocols used by France, UK, and Germany

Test France UK Germany
Dynamometer Test
NEDC Type approval test  
NEDC test with warm start  
NEDC test with variation of hood position, four-wheel dyno, altered shifting pattern    
NEDC but with slightly modified speed trace  
NEDC test in cold ambient conditions (10 °C)    
On-Road Test
NEDC speed trace
NEDC but with slightly modified speed trace    
RDE testing  

In all three cases, the government ministries overseeing the programs followed up the vehicle tests by asking manufacturers to explain the high emissions from their vehicles. But it does not appear that any of the ministries are conducting further, more customized, in-depth testing to examine the validity of the manufacturers’ claims. Only the UK and German reports give any detail as to how manufacturers are justifying their high in-use emissions.

The UK report accepts at face value the manufacturers' explanations for higher emissions under various circumstances. For example, the UK testing found that many vehicles had elevated emissions during a hot-start NEDC test. Manufacturers explain this as a product of higher engine-out NOx emissions from a warm engine than from a cold one. This explanation is not credible, as emissions control catalysts are more effective at higher exhaust temperatures, and a higher percentage of exhaust gases can also be recirculated in a warm engine to reduce NOx. But the only way to confirm or refute the manufacturers' explanation definitively would be to measure a number of variables—things like exhaust gas recirculation (EGR) rate, engine-out NOx, tailpipe NOx, exhaust temperature, coolant temperature, and inlet manifold temperature—and do a correlation between the cold-start and hot-start runs to determine if the emissions control system was reducing efficiency once it detected it was not being tested over a cold-start cycle. Further, you would have to determine which parameter was being sensed in order to let the engine control unit know when to modify the emission controls.

Both the UK and German ministries accepted, to at least some degree, manufacturers' justification for high emissions during on-road testing, namely that the tests were performed at low ambient temperatures (10 ˚C to 17 ˚C, depending on the manufacturer) in which NOx emissions controls must be relaxed to protect the vehicle’s engine. Given the lack of guidance and clarity in the EU regulation, as discussed in the KBA report (pp. 122–125), this may prove to be a permissible exception to the ban on defeat devices in the regulation, but it is a technically dubious claim. It also evades the contradictory evidence that a number of vehicles are in fact able to meet emissions limits in cold temperatures. It does not appear that the German or UK teams attempted to determine whether other defeat devices, in addition to the ambient-temperature (or "thermo-window") defeat device or the VW-style cycle detection defeat device, were in use in the vehicles they tested. Because nearly all the on-road testing in the three test programs was conducted during the winter months, none of the testing would have been able to differentiate between indications that a thermo-window defeat device was in use and indications that a vehicle was employing any other defeat device which senses real-world driving.

The UK testing in particular focused on trying to find a defeat device that works in an identical fashion to the VW defeat device. This unfortunately, and unnecessarily, restricted the scope of that investigation, because there are so many other parameters that a defeat device could be sensing—steering wheel position, usage climate control, rear wheels moving, GPS sensor, and cold start detection, to name just a few.

The potential number and variety of defeat devices is limited only by the creativity of the engineers who design them. It would be a mistake to rely solely on a defeat device testing program that screened only for known types, which in the present case number only three (by our count): The VW-type defeat device, which recognizes when the car is being driven on the test cycle; a thermo-window defeat device that senses when the ambient temperature is outside the conditions specified for the type-approval test “cold”; and a hot-restart defeat device based either upon a timer or on the temperature of the engine when it is restarted. In light of the suspicious NOx emissions behavior (small changes in the test protocol resulting in dramatic changes in emissions) documented by these testing programs, it seems likely that defeat devices are being employed in Europe that have yet to be detected. Further testing is needed, starting with the worst-offending vehicles from the UK, German, and French testing programs, if the goal is to independently determine the actual cause of the high emissions.