The new data was recorded in 2022, using fuel consumption monitors installed on-board vehicles first registered in the EU in 2021 and 2022, and reported to the Commission in 2023. Compared to the previous publication, the data for cars now originates from a broader sample of 2.2 million vehicles registered in 2021 and 740 000 vehicles registered in 2022.
The analysis confirms that the real-world CO2 emissions and fuel consumption from diesel and petrol cars on the road are, on average, around 20% higher than indicated by the official values from the standardised WLTP type-approval test used for regulatory purposes. This discrepancy is in line with what the European Commission had anticipated. The analysis also showed that the gap observed for cars is similar across the two registration years monitored so far.
The new data for plug-in hybrid electric cars registered in 2021 confirms the gap observed in the first reporting year, with real-world values on average 3.5 times higher than laboratory values. The new findings confirm that these vehicles are currently not realising their potential, largely because they continue not to be charged and driven electrically as frequently as anticipated. This was also the case for plug-in hybrid vehicles registered in 2022, despite having larger batteries and longer electric ranges.
As the dataset continues to expand over the coming years and its representativeness of the evolving fleet improves, it will be possible to determine how the gap evolves and assess whether further steps are necessary to prevent it from growing.
Background
Since January 2021, new cars and vans that can run on liquid fuels must be equipped with approved on-board fuel consumption monitoring (OBFCM) devices before they can be placed on the EU market. These devices record the vehicles’ fuel or energy consumption and the total distance driven.
This on-board data must be collected by manufacturers – either through direct data transfer from the vehicle to the manufacturer over-the-air, or when vehicles are brought in for repairs or servicing – and sent annually to the Commission. Since 2023, Member States have also begun collecting these data during periodical technical inspections and will report this data as of 2024.
Data
The dataset published today compiles data collected in 2022 and reported to the Commission in 2023.
For cars, the reported dataset includes 26.7% of cars first registered in the EU in 2021 and 18.6% of cars first registered in the EU in 2022. The data coverage has more than doubled for 2021 registrations compared to the previous year, when 10.6% of cars were reported.
The dataset for vans remains more limited, covering 2% of vans first registered in the EU in 2021 and 18% of vans first registered in the EU in 2022. This broader coverage for vans first registered in the EU in 2022 results from the share of vans required to be fitted with an OBFCM device increasing in 2022. While the coverage has improved for vans, the sample primarily covers heavier diesel vehicles.
Overall, data coverage has improved compared to the first year. However, this coverage still varies widely per manufacturer and remains below expectations for several manufacturers.
Results for conventional vehicles
In this second reporting year for cars first registered in 2021, the average gap observed between real-world and official type-approval average CO2 emissions and fuel consumption was 19.8% (1.2 l/100km or 28 g CO2/km) for petrol cars, and 18.2% (1.1 l/100km or 28 g CO2/km) for diesel cars. While findings for diesel cars align with those of the first reporting year, the average gap for petrol cars is somewhat reduced from the previously reported 23.4%.
The new data appears more reliable due to the improved manufacturer coverage and increased mileage. The vehicle sample was also more representative of the average petrol car fleet, being less biased towards heavier vehicles.
For cars first registered in 2022, the data shows an average gap of 21.1% (1.3 l/100km or 29 g CO2/km) for petrol cars, and 17.1% (1.0 l/100km or 26 g CO2/km) for diesel cars.
The average gaps observed for cars are compatible with the assumptions made for the 2021 and 2022 gaps in the Commission’s Impact Assessment, which underpinned the revision of the vehicle CO2 standards. Such gaps were anticipated as real-world emissions are influenced by factors which are not fully replicable in laboratory tests, such as traffic and road conditions, landscape, ambient temperature, use of air-conditioning and on-board electronics, and driver behaviour.
For vans, the sample data remains insufficient to draw conclusions. In the sample of 85,000 diesel vans first registered in 2022 (representing 9% of total 2022 registrations for this category, with a significant bias towards heavier vans), a preliminary real-world gap of around 13% was estimated.
Results for plug-in hybrid electric vehicles
For the 320,000 plug-in hybrid electric vehicles (PHEVs) first registered in 2021, included in the dataset of the second reporting year, the real-world CO2 emissions were on average 3.5 times (99 g CO2/km) higher than the type-approval values. This large gap aligns with the findings of the first reporting year.
For PHEVs first registered in 2022, the sample of 117,000 vehicles indicates that the real-world gap remained stable in absolute terms (from 99 to 102 g CO2/km), but increased significantly in relative terms (from 249% to 314%) compared to the sample of PHEVs first registered in 2021. Larger batteries and longer electric ranges in more recent PHEVs resulted in lower official emission values due to the assumption of more electric driving when calculating test results. However, the data suggests that this increased range has so far not resulted in a higher share of electric driving on the road, confirming that on average PHEVs are not being charged and driven fully electrically as frequently as anticipated.
The Commission has already introduced changes to the utility factor calculation – namely, the expected share of distance driven electrically based on the electric range – which is used to determine the CO2 emissions during the official test procedure. These changes will apply in two stages: first as of 2025, with a further adjustment in 2027. The Commission will continue to monitor the need to further adjust this utility factor based on real-world data.
Data publication
An overview of the findings, including data on each manufacturer’s new fleet of vehicles, was published by the European Environment Agency.
Read more: