Intended to address the shortfalls of the discredited New European Driving Cycle (NEDC), the Worldwide Harmonised Light Vehicle Test Procedure (WLTP) is supposed to bring CO₂ emissions figures in line with reality. Nick Molden, CEO of Emissions Analytics, has found it’s not working as intended.
Vehicle emissions testing across the EU is currently in a messy switchover phase.
The NEDC dates back to the early 1990s and was part of the background to Dieselgate, failing increasingly over time to measure real emissions as manufacturers optimized vehicles, whether legitimately or illegitimately, to game test outcomes.
Recognizing the failure of NEDC, the EU has, since 2017, begun a new laboratory test called WLTP, which is much closer to how drivers use their vehicles.
The problem is that while all cars sold from September 2018 must be WLTP-certified, the fleet average CO₂ figure – i.e. the number that is published for the vehicle – is still based on the old NEDC value as the full switchover to WLTP for published CO₂ figures will not take place until 2021. At the same time, the fuel economy figures published in the showroom are already based on the new WLTP system.
Confused yet? It gets worse… A situation has now arisen whereby WLTP figures are being subjected to complex translation mechanisms to derive NEDC-equivalent values from the WLTP test. While the EU has provided a software package to model this translation, NEDC-equivalent WLTP figures can be derived from manufacturer-declared values or from actual dynamometer tests.
Unquestionably, this means the system is still open to manipulation, disguises inherent flaws in WLTP, and leads to consumer confusion: hardly what you might expect almost four years on from Dieselgate.
However, diesel is no longer the problem. While we (Emissions Analytics) have observed that the real-world emissions for diesel cars are now close to being in line with WLTP results, the NEDC-equivalent WLTP CO₂ emissions of gasoline cars have fallen despite the test getting tougher.
We’ve discovered this because we have the EQUA Index, which we created in 2011 to provide the market with reliable information, and which has since collected real-world data on over 2,000 vehicles. It’s a database that allows for effective comparison between old NEDC figures, new adjusted NEDC figures, and WLTP results. What it has discovered raises serious concerns over the effectiveness of WLTP in policing the automotive contribution to climate change.
The gap between EQUA real-world data and the adjusted NEDC values for diesels has shrunk to the point where they are essentially the same. The gap with gasoline, however, has increased to 24%.
CO₂ values for diesel cars rise, as would be expected for a tougher test, between NEDC and WLTP. The average WLTP result of 175 g/km CO₂ is very close to the Emissions Analytics’ real-world average figure of 173g/km.
Gasoline cars on the other hand fell, counterintuitively, under the new test. The WLTP average figure was 151g/km compared with 152g/km under the NEDC. The NEDC-equivalent value – the one used to determine fleet averages – was even lower at 133g/km. All these values still lie far below the real-world figure of 185g/km measured by Emissions Analytics.
So, what we are left with is a significant discrepancy in relative certification values for diesel and gasoline vehicles, despite no change in real-world performance as measured by the EQUA Index over the last few years.
A potential explanation is that focus and expertise have potentially been put into optimizing gasoline vehicles on the new WLTP cycle, driving the official CO₂ down despite this not translating into better real-world performance. With the increased market share of gasoline vehicles post-Dieselgate, reduced CO₂ emissions from these vehicles makes the most impact in meeting fleet average emissions targets.
This would also make WLTP questionable as a source for gathering actionable data, much like NEDC that preceded it.
It is also worth pointing out that the EQUA Index currently shows real-world emissions sitting well above the 130g/km fleet average target for CO₂, which suggests official data flatters how much CO₂ is being reduced overall.
While laboratory figures are a useful way to test vehicles, there is an important lack of validation and correlation with real-world use. Granted, repeatability suffers when testing on the road, but it limits the ability of manufacturers to optimize for the test.
Emissions Analytics maintains that the best solution is for a simple and comprehensible certification program, combining laboratory tests and on-road validation, analogous to the Real Driving Emissions regulations for nitrogen oxides. Real-world surveillance would add extra deterrence to manipulation.
Until such a situation becomes normal, we are going to continue to lack clarity on real-world emissions from an imperfect test regime caught in a switchover phase.
