Created by
Biagio CIUFFO Scientific Officer
European Commission - DG JRC, Directorate C - Energy, Transport and Climate
Policies and Targets
Technology and Standards
Transport and Mobility

The European Commission needs to amend European Regulations 443/2009 1 and 510/2011 2 due to the introduction of the Worldwide harmonized Light-duty vehicles Test Procedure (WLTP) in the European type-approval of light duty vehicles. The new test-procedure, developed at the United Nations Economic Cooperation for Europe (UNECE) level in the period 2009-2015 3, was planned to replace in Europe the current type-approval procedure, based on the outdated NEDC (New European Driving Cycle) procedure. In the recent years, indeed, several sources have outlined the incapability of the NEDC to provide reliable estimations of fuel economy and emissions, showing the existence of a widening gap between type-approval values and those experienced by the vehicle users in their daily driving (4,5). For this reason the European Commission (EC) had already targeted to have a new and more representative test procedure in place as of 2017.

From a political perspective, achieving this objective is more challenging than it may seem, as it requires the amendment of the CO2 emission targets for light duty vehicles defined in the two aforementioned regulations. These targets have been defined to support European Members States meeting their binding CO2 reduction targets defined in the framework of the Europe 2020 strategy (setting an overall target for Europe to reduce greenhouse gas emissions by 20% in 2020 from 1990 levels, 6). The targets are assigned to vehicle manufacturers (OEMs in the following) and represent the maximum allowed average specific CO2 emissions (in g/km) for the vehicles registered in European fleet during an entire calendar year by that manufacturer. Non-compliant manufacturers have to pay an excess emissions premium for each car registered (7). Two targets are defined for 2015 and 2021, requiring manufacturers to achieve overall emission targets of 130 gCO2/km (2015) and 95 gCO2/km (2021). These emission figures are based on the current type-approval process and in order to meet them each OEM has chosen a certain strategy (leading for example to the introduction of hybrid, plug-in hybrid and electric vehicles way before they reached clear economic convenience). If, as already mentioned, the EC has the intention to introduce by 2017 a new test procedure in the vehicle type approval, a legal basis is needed to deal with the existing targets. This legal basis is included in the aforementioned regulations, which foresee the necessity to amend the targets due to the introduction of a new test procedure with the addition of "ensuring comparable stringency for manufacturers and vehicles of different utility". The legal interpretation of this additional precision is that if an OEM is able to reach its target on the current test procedure, it should be able to do so also on the new one. This interpretation has been strongly supported by the European Members States who have followed this process. In order to achieve this objective, new WLTP-based targets should have been designed knowing the perfect relationship between CO2 emissions under the two test procedures for each vehicle in the market in 2015 and in 2020, which was something impossible to achieve.

Supporting the EC on this problem, the authors proposed instead the following alternative solution: in 2017 WLTP is introduced in the vehicle type-approval. Vehicle tests are carried out following the new test procedure. A vehicle simulation model is available at that time to calculate the corresponding CO2 emissions from the previous procedure so that new WLTP-based targets are not necessary. In this way, the conditions to introduce the WLTP are all respected and the following benefits hold:

  • As of 2017 emission limits for regulated pollutants (NOx, CO and PM/PN) are measured using a more realistic test procedure (for them, no changes in the limits are necessary) and this would provide an immediate benefit for air quality;
  • although CO2 targets are still assessed with respect to the old test procedure, more realistic CO2 figures will also be available. Also in this case this will represent an immediate benefit for consumer information;
  • post-2020 CO2 targets can be directly set on the new test procedure. Without the WLTP in place, the old test procedure shall be used instead;
  • using a simulation model to evaluate CO2 emissions from a vehicle have the benefit to make the entire process more deterministic preventing an OEM to take excessive advantage from the so-called "flexibilities" in the old test procedure (4).

For all these reasons, the EC supported the proposal by the authors and CO2MPAS (CO2 Model for PAssenger and commercial vehicles Simulation) is the model developed for this purpose. 

CO2MPAS follows a hybrid approach between a more general-purpose emission model (which attribute average performance indicators, namely emission factors, to the various vehicle types and classes) and a fully developed vehicle simulation model (which usually use extensive component measurements to map emissions in order to later reproduce them for diverse engine states and vehicle operating conditions). The key design conditions of the model are:

  • Capacity to capture the effect of specific fuel/energy consumption reduction technologies, when individually applied or in combination, at vehicle level with accuracy in the order of 2-3% as compared to measurements;
  • Ability to perform accurate simulation of vehicle fuel consumption/CO2 emissions under different operating conditions using mainly physical models and by limiting statistical factors as much as possible;
  • Limit the number of inputs to the vehicle characteristics already available during the vehicle type approval.

Core of the simulation module is a physical model based on standard vehicle longitudinal dynamics and energy consumption simulation. Initial investigations indicated that the most important factors affecting CO2 emissions over given driving condition are:

  • Accurate calculation of power;
  • Driving behaviour (gear-shifting, acceleration patterns);
  • Powertrain operation and efficiency;
  • Cold start – temperature conditions;
  • Controls of secondary systems – power sources and/or sinks.

The model has been thoroughly validated with respect to the results achieved using a different and more sophisticated vehicle simulation model and with respect to the results of real measurements. Figure 1 presents a summary of the validation of the model with respect to real vehicle measurements. The model is always able to predict CO2 emissions with an error lower than 5% and in more than 70% of the cases the model is able to estimated measured CO2 emissions with an error lower than 3%. 

Figure 1. Comparison of measured and simulated CO2 on 21 vehicles using CO2MPAS. Blue bars represent test results, while red bars represent simulated values. In the horizontal axis the following abbreviations are used: AMT (Automotized Manual Transmission), AT (Automatic Transmission), MT (Manual Transmission)

The model and the procedure to apply it have been developed with the large stakeholder involvement. Technical experts from Member States, vehicle manufacturers and non-governmental organizations have supported the European Commission in the developing the approach finally adopted. As a final result, the Implementing Regulation XXX has been approved by the European Council :::. In the regulation, for the first time, explicit reference to the use a simulation model for regulatory purposes has been made, as well as to the boundary conditions to ensure its correct application. Without this regulation the WLTP could not have been introduced in the European Type-Approval process with the consequent effect of at least one decade of delay in the roadmap defined by the European Commission to have a less carbon intensive transportation sector. In the near future, a second case of the use of simulation for regulatory purposes in vehicles-related legislations will be presented. The VECTO model, also developed by the European Commission in collaboration with the Technical University of Graz will be used to certify the fuel consumption of Heavy Duty Vehicles.

Both cases represent a clear example where science, used in a fully participatory way, can significantly help the policy in achieving objectives that, otherwise, with the only logic of the political compromise, it would have been very difficult to achieve.

The CO2MPAS model has been developed under the EUPL scheme. The software as well as the source code and the instructions to use it can be found and downloaded from the following repository.
This site is in beta version - some features may be unstable. Please report issues to the E3P Webmaster using the contact form.