CO2 emission standards for cars and vans: how to make the switch to a WLTP-based compliance system a success

My last blog post dealt with the changeover of the CO₂ regulation for light duty vehicles from the New European Driving Cycle (NEDC) to the Worldwide Harmonized Light Vehicles Test Procedure (WLTP) and its potential weaknesses with regard to the effectiveness in terms of real-world CO₂ emission reduction. By switching from emission target levels in absolute numbers to relative reduction targets with respect to the WLTP target level in 2021, the EU Commission’s proposal has unintentionally created a huge incentive for the carmakers to achieve high WLTP to NEDC CO₂ emission value ratios by 2020. The higher the ratio between the WLTP and the NEDC in 2020, the higher the WLTP target level from 2021 onwards will be. My last post also showed the existence of a regulatory loophole and it mentioned the possibility of vehicle optimization as well as portfolio management as potential means to inflating the WLTP targets levels beyond 2020 to disproportionately high CO₂ emission values. Last but not least, I argued that there is the potential to address these current shortcomings of the transition from the NEDC to the WLTP based compliance system. And here we are. Our ideas on how to deal with this transition are the topic of this blog post.

Addressing potentially inflated WLTP targets levels beyond 2020: what is important for potential measures against unjustified high WLTP CO2 emission targets

Before sharing our thoughts on potential regulatory improvements, I would like to state criteria which we had in mind when discussing potential options for limiting disproportionately high WLTP CO₂ emission targets beyond 2020.

1. The regulatory adjustments should prevent the unjustified inflation of WLTP emission targets as much as possible. Thus, existing procedures and very quickly applicable solutions are preferred. This prerequisite eliminates adjustments which require new type-approval procedures and tests since they require “too much” time being developed and implemented. Additionally, applying existing procedures usually minimizes the administrative burden and reduces the possibility of unknown weaknesses of newly developed procedures.
2. The regulatory adjustments should correct emission target levels as soon as unjustified high WLTP emission targets levels are observed within the monitoring data. It might require adapting the number and the type of monitored data during the certification and type-approval process of new vehicles.
3. The regulatory adjustments should be as vehicle- and manufacturer-specific as possible. The transition process from the NEDC to the WLTP CO₂ emission standards is based on vehicle-specific – and thus on manufacturer-specific – correlation factors. Manufacturers might exploit the weaknesses of the transition process at a different level and different vehicle portfolios could result in varying translation factors [1]. As a consequence, the WLTP emission target correction – if required – should apply at a manufacturer-specific level since it would take the manufacturer-specific situation into account. It would also reduce the competitive need of each carmaker to engage in inflating the WLTP emission values. (see the last blog post for the explanation of why the competition between carmakers creates an incentive for each manufacturer to inflate the WLTP emission targets).

Keeping these three criteria in mind, here are our ideas:

Addressing potentially inflated WLTP targets levels beyond 2020: closing the regulatory loophole

The most straightforward approach to combating “too” high WLTP CO₂ emission targets beyond 2020 is closing the regulatory loophole of the correlation procedure between the WLTP and the NEDC emission values. In the recent blog post, I explained that there is no link between the declared WLTP and the declared NEDC type-approval CO₂ emission value of a vehicle. The declared WLTP value can be set to any emission value above the measured emissions during the WLTP type-approval procedure. Thus, the manufacturers can virtually choose their own WLTP to NEDC ratio in 2020. Eliminating the possibility to freely inflate the declared WLTP CO₂ emission value would close the regulatory loophole immediately. So, let’s think about how to do it!

Figure 1:   Schematic representation of how amending the correlation regulation between WLTP and NEDC CO2 emission values could close the regulatory loophole in the changeover from a NEDC to a WLTP based compliance system.

Source: Own illustration

In my opinion, the best option is amending the regulation for the correlation process between the WLTP and the NEDC emission value (for cars EU/2017/1153 and for vans EU/2017/1152). Currently, the correlation regulation sets a lower limit for the declared NEDC emission of a vehicle as the declared value must not be significantly smaller than the NEDC equivalent CO₂ values from the correlation tool CO₂MPAS. The regulation also implements a verification procedure including a sanctioning mechanism for non-compliance with the lower NEDC emission bound (see our paper for more information about it). An upper bound for the declared WLTP emission value could be added to the correlation procedure by a similar mechanism: for verification, the declared WLTP CO₂ emission value shall not be higher than the measured one of the WLTP type-approval procedure (see Figure 1). A measurement tolerance value should also be applied to consider the minimal differences in the testing environments. A link between both the declared WLTP (the upper limit would be set through the amendment) and the declared NEDC emissions (the lower limit has already been part of the correlation regulation) would be established by adding this prerequisite to the correlation regulation. The existing verification procedure including the sanctioning mechanism for the lower bound of the NEDC emission values could be extended to also cover the upper limit of the declared WLTP emission value. That’s it! The regulatory loophole would be closed. There is another option to deal with the regulatory loophole if the amendment of the correlation regulation appears to be inappropriate: a WLTP emission target correction mechanism which is based on the gap between the declared and the measured WLTP emission values could be added to the CO₂ emission standards for cars and vans. It would facilitate the equivalent impact as the above proposed amendment of the correlation process (see Figure 2). In my opinion, this is the second best option to closing the regulatory loophole of the transition from NEDC to WLTP since it would result in constant adjustments of the carmakers’ targets and would make the regulation less transparent to non-experts. Additionally, the target adjustment would always lag behind since the monitoring data would become available only in the year after the registration of the vehicles. Monitoring of the difference between the declared and the measured WLTP CO₂ emission value is the prerequisite for a correction mechanism and would have to be added to the WLTP type-approval and the WLTP-NEDC correlation process.

Figure 2:  Schematic representation of how a WLTP target correction mechanism could close the regulatory loophole of the changeover from a NEDC to a WLTP-based compliance system.

Source: Own illustration

Both options for addressing the regulatory loophole of the transition from NEDC to WLTP seem pretty straightforward and easy to implement. What do you think? Should one of the options be implemented? Is there any downside to the suggestions? Actually yes, there are two downsides. Firstly, the transition to the WLTP already began in September 2017 and the current certification procedure has not foreseen the monitoring of the measured WLTP CO₂ emission values. Thus, the measured CO₂ emission values of the vehicles which have already been certified under the WLTP are not available. This would require recertification of these vehicles or the changes of the regulation would only be effective for vehicles which are type-approved after the changes became effective. And the second downside? Both options “only” address the regulatory loophole of the transition from NEDC to WLTP: vehicle CO₂ emission optimization and portfolio management in 2020 would still be possible to maximize the WLTP emission target level from 2021 onwards.

Addressing potentially inflated WLTP targets levels beyond 2020: continuing the monitoring of the NEDC emission values beyond 2020 might help

So, let’s think about how to address other issues than the regulatory loophole and how to react comprehensively to the incentive of reaching high WLTP to NEDC CO₂ emission ratios by 2020. Our answer to this question might be a little surprising to you: continuing the monitoring of the NEDC CO₂ emissions beyond 2020 might be a very helpful instrument to increase the effectiveness of the transition from the NEDC to the WLTP. But why should the continuation of a rather imprecise test procedure support the effectiveness of the CO₂ emission standards? Here are our thoughts on that.

Figure 3: Schematic representation of how a WLTP target correction mechanism based on an appropriated measure of the relation between the WLTP and NEDC CO2 emission values could address the issue of potentially disproportionately high WLTP CO2 emission targets beyond 2020. Both cases (target correction and no target correction) are shown below.

Source: Own illustration

The manufacturers achieve high WLTP CO₂ emission targets beyond 2020 by establishing high WLTP to NEDC CO₂ emission ratios for their new vehicle fleet in 2020. In order to understand if the emission ratio of 2020 is reasonable, it could and should be monitored beyond 2020[2]: a rather constant relation between the two values indicates that the “new” WLTP emission target by 2021 is an appropriate starting point for the period up to 2030; the opposite holds true for decreasing ratio values. Eventually, a WLTP CO₂ emission target correction mechanism could be applied to adjust the targets based on the findings of the WLTP and NEDC monitoring data beyond 2020 and the development of the relation between WLTP and NEDC CO₂ emission values (see Figure 3). Thus, this approach would meet the principle of “comparable stringency” by 2020 and would reduce the risk of making the changeover to a WLTP based compliance system rather ineffective in terms of CO₂ emission reduction. In my opinion, there is another plus to this approach. Indirectly, the WLTP real-world emission gap which we will probably see increase after 2021 would be roughly monitored and the target correction mechanism would roughly limit the impact on smaller emission reduction due to the rising real-world emission gap. The reason for this is the link between the relation of the WLTP to NEDC CO₂ emission values and the increasing WLTP real-world emission gap. Until 2020, the CO₂ emission standard compliance will be monitored in NEDC values, and a high WLTP to NEDC CO₂ emission ratio will be beneficial for manufacturers in 2020. However, beyond 2021 the CO₂ emission compliance will be based on WLTP emission values. And thus, I’m expecting a switch from vehicle and portfolio optimization towards low NEDC and high WLTP CO₂ emission values up until 2020 to CO₂ emission optimization towards low WLTP emissions from 2021 onwards. And the result of this switch in emission optimization should be the following if no target correction mechanism is applied: the measure for the relation between the WLTP and NEDC emission values will decrease after 2020 and the WLTP real-world emission gap will increase due to CO₂ emission optimization towards the WLTP. As a result, both issues would be addressed by tracking the  relation between the WLTP and the NEDC emission values and how it develops over time. And the best part? All the required data for this approach is already available and monitored. It would only require extending NEDC emission monitoring beyond 2020 until new testing and monitoring procedures (e.g. onboard monitoring devices, PEMS tests) allow more precise regulations on the real-world effectiveness of the WLTP CO₂ emission reduction targets. Of course, the design of an appropriate correction mechanism would also be needed.

Addressing potentially inflated WLTP targets levels beyond 2020: let’s bring it together

In the beginning of this post, I mentioned three criteria for evaluating concepts for dealing with disproportionately inflated WLTP CO₂ emission targets beyond 2020. In my opinion, our ideas meet these criteria. All of them are applicable at the vehicle-specific (amendment of the correlation regulation) or manufacturer-specific level (both suggestions for a target correction mechanism). Check! And each of them uses data which is available today, but which is partially not monitored or would have to be applied differently. No completely new procedures would be needed, and existing regulations would “only” have to be updated. Thus, the implementation of those concepts should be possible rather quickly. Check! A long blog post again! I assume that most readers are more interested in thinking about the appropriate ambition level of the CO₂ emission standards and incentives for zero emission vehicles. It is easier to grasp and communicate. I will cover those topics in upcoming blog posts of this series on CO₂ regulation. However, the reason why I have started the series with these rather technical issues is the lack of attention on the topic. And the potential to negate the CO₂ emission reductions achieved by introducing a WLTP based compliance system is huge. And thus, its discussion is particularly important. In my opinion, an amendment of the correlation regulation would work well to close the regulatory loophole of the transition from a NEDC to a WLTP based regulation system. And it would be important to implement a WLTP target correction mechanism for a potentially decreasing gap between WLTP to NEDC CO₂ emission values beyond 2020. What is your opinion? Do you have other concepts in mind? And is it really an important issue to be considered more than it is now? I look forward to discussing these issues with you after my summer vacation :-).

[1] Different technical characteristics (e.g. petrol and diesel cars) might result in varying translation factors (see Tsiakmakis, S. et al. (2017). A simulation-based methodology for quantifying European passenger car fleet CO₂ emissions. Applied Energy, (199), pp. 447–465 for more information on that). [2] Research on the detailed design of a good measure for the relation between WLTP and NEDC (e.g. the difference, the ratio, a mix of difference and ratio between both values) emissions would be needed.

Peter Kasten is a senior researcher in the Resources and Transport division. He has joined the Institute in 2010 and is leading the work on the interactions between the transport and the electricity sector and is advising the German Ministry of the Environment on light-duty vehicle CO₂ emission targets.

To Peter Kasten's first blog post: "CO2 emission standards for cars and vans: the status quo and the Commission’s proposal"

To Peter Kasten's second blog post: "

CO2 emission standards for cars and vans: the difficult switch from one test procedure (NEDC) to another (WLTP)"