At the ICCT we have been analyzing the growing gap between official and real-world vehicle fuel consumption and CO2 emissions of passenger cars in Europe for a few years now. We showed for the first time in 2012 that on-road CO2 emissions and fuel consumption of new cars were not as low as the laboratory tests conducted for type approval indicated. At that time we analyzed data for 28,000 vehicles from two sources. For our most recent analysis, released this past September, we extended the scope to more than 600,000 vehicles from eleven data sources. But the overall trends did not change – the gap keeps growing, at an ever-increasing pace.
Observing and analyzing the growing gap is certainly important to create and maintain awareness of the problem that we are facing. At the same time, it is important to better understand why the gap keeps growing and what we can do about it. Earlier this year, we prepared a report for the UK Committee on Climate Change, detailing a long list of loopholes in the current European vehicle emissions testing scheme and analyzing how these are increasingly being exploited by vehicle manufacturers.
Around the same time, I began a long trek into the jungle of the European new-vehicle type-approval system. I call it “jungle” because it is difficult to see through — difficult to understand what happens before a new vehicle model receives its type-approval certificate and can be offered to customers, as well as to understand what happens (or does not happen) in terms of emissions testing, once the vehicle is for sale on the market. I quickly found myself deep in a thick underbrush of technical documents defining how vehicle emissions testing ought to be carried out in the EU, with most of these originally developed many years ago and revised and amended numerous times since.
Once I had a better understanding of the principles of vehicle emissions testing in the EU, I asked my colleague John German for help. He is very familiar with how vehicles are tested in the U.S. and so we pooled the knowledge we had collected on both sides of the Atlantic and prepared the comparative overview chart below.
What I find most interesting about this comparison is that the actual vehicle testing process in the laboratory is not that different. Here in Europe we use the New European Driving Cycle (NEDC) as the basis, and in the U.S. they use the Federal Test Procedure (FTP) and the highway cycle (or, in the case of air pollutant emissions testing, a combination of five different test cycles).
But what is very different in the U.S. is that the regulator, the U.S. Environmental Protection Agency (EPA), carries out confirmatory testing on about 15 percent of pre-production test vehicles and also has a robust in-use vehicle testing program for air pollutant emissions. So when a vehicle manufacturer in the U.S. measures the emissions of its vehicle models and forwards the results to EPA, it knows that some of its vehicle models will be subject to independent re-testing by EPA, with repercussions if the manufacturer’s results are statistically different from EPA's testing.
In the EU, on the other hand, this kind of re-testing does not occur. In practice, often what happens is that manufacturers test their vehicle models in their own laboratories and pay an accredited technical service company to witness the testing. The measurement results are then forwarded to the type-approval authority, but are rarely questioned. In fact, type-approval authorities in the EU often do not even have facilities where they could carry out confirmatory tests. Furthermore, the manufacturer is not obliged to get its vehicle models type-approved in a specific member state. Instead, it can choose a member state, and thereby a type-approval authority, that suits it best, creating a perverse incentive for national type-approval authorities to relax their requirements in order to attract more “customers.” Once a type-approval certificate is issued in one member state it has to be recognized by all the other EU member states as well.
The same is true for the determination of coast-down values, which are used to set up the equipment for vehicle type-approval emissions tests so that it imitates road-load forces like tire friction and aerodynamic drag. These values are absolutely crucial to getting laboratory test results that correspond to actual on-road performance. And it's the same story: in the EU, coast-down testing that determines those essential values for a given vehicle model is carried out or commissioned by the manufacturer, with no re-testing by the authorities. Worse, the coast-down testing results are kept strictly secret in the EU. Why? I don’t know. But I do know that in the U.S., not only do the agencies carry out independent coast-down confirmatory testing, but anyone can download the official coast-down figures for each and every vehicle – not only those re-tested by EPA – from EPA's website. This allows research organizations like us to analyze and cross-check the manufacturers’ data, which is impossible to do in Europe.
I could go on like this ... In-use surveillance testing of vehicles on the road? In the EU it exists only for air pollutant emissions, but not for CO2 and fuel consumption. Again, it is the manufacturers carrying out the testing, with no independent re-testing foreseen by the European regulations. If an EU member state decides to carry out its own in-use surveillance testing (as Sweden and Germany have done in the past), then these results have no legal consequences. In the U.S., on the other hand, the regulatory agencies routinely conduct random and targeted re-tests of the manufacturers’ in-use surveillance results, and there have been numerous recalls and fines for non-compliance.
I don’t want to imply that the U.S. system of vehicle emissions testing is perfect and that the EU or any other government should simply copy-paste the U.S. approach. But it is quite remarkable for me how effective the U.S. authorities are when it comes to enforcing vehicle emission standards. When I started writing “The future of vehicle emissions testing,” it was still a few months before Volkswagen admitted having used illegal methods to lower their vehicles’ emission values. At that time I wrote that the growing gap, both for CO2 and NOx emissions, has the potential to undermine the credibility of individual manufacturers and even the entire auto industry. And I wrote that in order to really solve the underlying problem, the EU would need to go beyond just introducing a new test procedure.
Before I could even finish the report, I was overtaken by reality. At least one carmaker now has a serious credibility problem, if not the entire auto industry. And it has never been more obvious that introducing the WLTP and RDE test procedures by itself will fix only half of the problem (see here for an excellent summary of the latest developments around RDE). At least as important is a major revision of the European type-approval legislation, putting an end to today’s situation where a vehicle manufacturer can cherry-pick a member state to certify a vehicle and there is no validation on in-use vehicles.
What is ultimately needed is oversight of vehicle emissions testing by a European-wide authority, without any potential conflict of interest, as recently proposed by European Commissioner Elżbieta Bieńkowska and members of the European Parliament. This authority should also have the right and mission to re-test vehicles. And not just the “golden cars” that manufacturers specially prepare for laboratory testing, but normal vehicles, just as they are driven by normal customers under normal conditions. It is only through this combination of revising the technical aspects of how we test the emissions of new vehicles and modifying the legal framework that we can align official and real-world emission and fuel consumption figures. And it is the only way to avoid ugly surprises and scandals in the future.