Project Overview

How do traffic lights, road conditions, traffic congestion, and driving behavior influence motor vehicle emissions out in the real world? Drs. H. Christopher Frey and Nagui M. Rouphail are faculty members in the Department of Civil Engineering at North Carolina State University and co-principal investigators of a recently completed project that was aimed at answering these types of questions. The project team included two graduate students, Alper Unal and James Colyer at NC State. The project, titled "Emissions Reduction Through Better Traffic Management," was sponsored by the North Carolina Department of Transportation. The final report (6.35 MB) is available on-line.

The NCSU project was aimed at measuring real-world on-road vehicle emissions with a focus on characterizing the effect of traffic signal timing on emissions. For example, we were interesting in knowing how the timing and coordination of multiple traffic signals affects vehicle emissions, both at specific intersections and when driving along a traffic corridor (e.g., a stretch of roadway that has many signals). How important are driving modes such as deceleration, idling, and acceleration when stopping at, waiting for, or leaving a signalized intersection? Do emissions vary for different vehicles?

Examples of Results

In our study, we found that an effective signal timing and coordination plan can reduce real-world on-road emissions. We also observed that emissions are lower on average for uncongested versus congested traffic on a particular corridor. Emission rates on a mass per time basis are highest during vehicle acceleration, and lower by a factor of five to ten, on average, for idling. Emissions are influenced by the number of stops and the amount of delay time experienced at intersections. A key difference between this study and many other vehicle emissions projects is that our results are based upon real-world on-road measurements of vehicle tailpipe emissions on a second-by-second basis. These measurements were made during actual driving, under actual traffic conditions, at any location along a route, with different drivers, and at various times of the day, days of the week, and months of the year. In fact, we made over 1,200 one-way trips while collecting data for this study. Many other studies are based upon measurements made on a dynamometer in the laboratory, remote sensing measurements which are obtained at specific locations representing only a snap shot of emissions, or tunnel studies that are not representative of driving on many other types of roadway facilities, such as signalized corridors. The results we obtained were based upon measurements, not models.

Key Findings of the Study

Some of the key findings of the study were:

This study has established the feasibility of using on-board emissions measurements to collect real-world on-road tailpipe emissions data for carbon monoxide (CO), nitric oxide (NO), and hydrocarbons (HC).
Emission rates for different driving modes (acceleration, deceleration, idle, cruise) emission rates are for the most part stable for the same vehicle/driver combination.
Measured emission rates (on a gram per second basis) are highest during the acceleration driving mode and lowest for the idle driving mode. The difference between the two is typically a factor of five to ten on average, depending on the pollutant.
Measured emissions tend to increase with traffic congestion since there are more acceleration events, as observed on Chapel Hill Road.
Signal improvements such as coordination and retiming have resulted in lower real-world emissions on Walnut Street.
There are emissions "hotspots", which are locations of higher than average emissions, at specific locations on the corridors studied. These hotspots are typically associated with a signalized intersection.

Recommendations

The key recommendations based upon the study work are:

On-board emissions measurement methods should be used in future studies to improve knowledge of real-world on-road tailpipe emissions.
On-road measurement studies should be carefully designed taking into account well-defined study objectives.
Objectives for future studies should include but not be limited to empirical evaluation of Transportation Control Measures (TCMs) and Transportation Improvement Projects (TIPs) and plans; assessment of emissions hotspots; evaluation of driver behavior; validation of emission factor models;and development of public education tools to educate drivers about their role in pollution prevention.
Appropriate and thorough data screening, reduction, and analysis protocols should be used.
Variability and uncertainty should be accounted for in the study design and when making inferences based upon measured data.
State departments of transportation and others should use on-road emissions data to aid in the design and real-world evaluation of TCMs and TIPs.
US EPA should use on-board emissions measurements to validate its emission factor models and to develop new emission factor models.

Implications

In addition to the importance of traffic signal timing with respect to vehicle emissions, the work we are doing has many other implications and possible future applications. For example, the data we collect can be used to characterize actual on-road emissions, as opposed to having to rely solely on computer models to make such predictions. The most commonly available and widely used computer models are based upon data obtained in laboratories for standardized driving cycles (which are typically trips or segments of trips). In contrast, our work is based upon measurements taken under real world driving conditions, including effects of road grades, roadway design, traffic signal timing, traffic conditions, weather conditions, etc. Thus, our data are more representative of real world emissions than the predictions of computer models. An important need for on-road emissions estimates is to predict the impact of vehicle emissions on air quality. It is also important to understand the relationships between traffic flow and emissions in order to develop effective air quality management strategies.

Another important implication of our work, aside from the potential for improving air quality by developing a basis for improved traffic signal timing, is to develop the data needed to understand the relationship between roadway design and vehicle emissions.

A third implication of on-board emissions measurements is that it can be an important way to supplement or replace existing emissions inspection programs.

If you have any suggestions for how to use this data, please feel free to let us know. You can contact members of the project team via the Miscellaneous Links portion of the web site.

Guide to this Web Site

For an overview of the equipment that used in this project, please click here or on the "Measurement Instrument" link below. For some sample results of measurements that we have taken, please click here or on the "Example Measurements" link.

You can also obtain a copy of a presentation giving an overview of the methods, some example results of this project, and the project final report, from the "Presentations" link.

As we obtain results from this project that may be of general interest to the public, we will post them on the "Low Emissions Driving Tips" portion of the web site.

You can access other information related to this project via the "Miscellaneous Links" portion of the web site.

This web page was designed by H.C. Frey
All items on this web site are Copyright (c) 1999 by H.C. Frey and NC State University
If you wish to use any materials from this web site, such as for publication or presentation, please contact Dr. H. C. Frey at frey@eos.ncsu.edu.