Integrating a Macro Emission Model with a Macroscopic Traffic Model

Integrating a Macro Emission Model with a Macroscopic Traffic Model


Gerdien Klunder, TNO, Henk Taale, Rijkswaterstaat, Uilke Stelwagen, TNO


This paper presents a macro emission module for macroscopic traffic models to be used for assessment of ITS and traffic management. It especially focuses on emission estimates for different intersection types. It provides emission values for CO, CO2, HC, NOx, and PM10.


In order to be able to assess the emission effects of traffic management and ITS on the level of a city or a region, a general approach is to use a mesoscopic or macroscopic dynamic traffic model and then to use the effects on the traffic flow to estimate the emission effects with an additional traffic emission model.
Generally, two types of traffic emission models exist, namely microscopic emission models that estimate emissions based on detailed individual vehicle data (speed and acceleration) and macroscopic emission models that use macro averaged emission factors and macro (aggregated) traffic data. For macroscopic traffic models, only a macroscopic emission estimation approach can be applied since detailed individual vehicle data is not available in these models. However, in order to incorporate the effects of changes in dynamic behavior, routing or traffic management strategies in a realistic and accurate way, a more detailed approach is needed. Unfortunately there are no emission factors available that can differentiate in any realistic nor accurate way for vehicle dynamics induced by different intersection types and traffic intensity variations. In order to derive such emission relations, detailed information is needed about traffic flow for different intersection types and relations should be found between macroscopic traffic variables and resulting emissions on a macroscopic scale. This is a challenge which has not been tackled before. Therefore a macroscopic emission module has been developed, which is explained in this paper.

The Ministry of Infrastructure and the Environment of the Netherlands has developed a macroscopic traffic model called the ‘Regional Traffic Management Explorer’ (or Regionale BenuttingsVerkenner, RBV). This model has proven to be an important tool in supporting regional planning and evaluation processes for traffic management. The core is a dynamic traffic assignment model called MARPLE (Model for Assignment and Regional Policy Evaluation). MARPLE can be used to determine the effects of traffic management measures such as traffic signal control, peak hour lanes, speed measures and buffers on a macroscopic scale. With the macro emission module, emissions of CO, CO2, HC, NOx, and PM10 are added to the output of MARPLE, for each link and time interval (normally in the order of 10-15 minutes).

In order to find relations between macroscopic traffic variables and macroscopic emissions, traffic micro simulations were performed for different types of intersections, with the micro simulation model VISSIM. Next, emissions were calculated with the TNO emission model VERSIT+ with the individual vehicle data from VISSIM (speed and acceleration) as input on a per second base. A thorough investigation revealed that only the mean speed averaged over all vehicles in the intersection traffic network showed a clear relationship with CO2 and other emissions (CO, HC, NOx, and PM10). This approach led to sets of emission rate curves (emissions in gram/second) as a function of the mean speed. Per intersection type and speed limit a set of size dependent curves was derived and these macro emission curve sets are used in the so-called macro emission module.
Challenges for this approach were how to cope with differences in links leading towards or away from an intersection, emissions on the intersection space itself, different link lengths and aggregation intervals. In the full paper it will be explained how these issues were tackled.

For the validation of the emission module, three types of comparisons have been made:
1) As a first verification the total emission calculated directly from the VISSIM/VERSIT+ microsimulations, was compared to the total estimated emission calculated from the derived macro relations applied to the VISSIM mean speeds. This comparison proved to be very acceptable with a relative difference of less than 0.1 % using 10 minutes averaged data.
2) Mean speeds on a macroscopic scale derived from the VISSIM microsimulations were compared to the macroscopic speeds from the RBV for the same intersection types and traffic settings. This showed larger differences which can be reduced by better tuning of MARPLE.
3) Emissions based on microscopic traffic data from the VISSIM/VERSIT+ microsimulations compared to emissions based on macroscopic data from the RBV and the macro emission relations from the macro emission module. In case of proper calibration, the results were satisfactory.

In the full paper, the emission module and validation results will be presented in more detail, as well as some conclusions, recommendations for further research and future outlook.


Association for European Transport