Civil Engineering ETDs

Publication Date

Spring 5-15-2021


Most roadway transportation modes still run on petroleum fuels. Petroleum fuels are major sources of air pollutants, including hydrocarbons (HC), carbon monoxide (CO), carbon dioxide (CO2), nitrogen oxides (NO and NO2, which are together called NOx), Volatile Organic Compounds (VOCs), and particulate matter of size(PM10) and2 and BC). Mitigating pollution is a serious issue, and therefore, it is the engineers' and planners' responsibility to study and reduce emissions as much as possible. Emission levels are generally high at street intersections, which are places where vehicles tend to accelerate, decelerate, and idle. Emission levels at intersections are a function of many different variables, including traffic signal timing and the mix of vehicle types.

This thesis explores how energy consumption levels, emission rates, and emission levels of five pollutants (HC, CO, NOx, CO2, and PM2.5) vary with cycle length, operating modes, and vehicle types at an isolated, two-phase signalized intersection. This study is based on VISSIM micro-simulation, MOVES emission models, and linear regression analysis (with emission rates as a dependent variable and cycle lengths, delays, and stopping rates as independent variables). A hypothetical intersection is chosen such that the vehicle distribution cars 90%, bus 5% and trucks 5%. The traffic volume is chosen such that V/C ratio is greater than 1. Therefore, a high emission is expected in the intersection and the intersection is reasonable to be optimized for the emission.

A key finding is there is a different optimal range of cycle length for each pollutant: 70s-110s, 100s-120s, 70s-120, 70s -120s, and 60s - 120s for HC, CO, NOx, CO2, and PM2.5, respectively. One of the reasons for different optimal ranges of cycle length for each pollutant is that different vehicle types use different fuel types, with each fuel having a different emission rate for each pollutant. The study also found that even though the proportion of the time travelled by vehicle while accelerating is less, it contributes to the highest proportion of total emissions than travelling in any other operating modes (cruise, decelerating and idle). Therefore, controlling the proportion of travel time in acceleration could help in controlling the amount of emission. The final optimal range of cycle lengths is chosen based on different scenarios. For an intersection in area with a lot of bus services and pedestrians/bikes, like downtown, the optimal cycle range is found to be 70s to 90s. Similarly, for an intersection in area with a large number of trucks and no bikes/pedestrians, like an industrial area, the optimal cycle length is found to be 90s to 110s. And for intersection in suburbs, where there are only cars and no (or few) bikes/pedestrians, buses, and trucks the optima range of cycle length is found to be 80s to 100s.

Finally, the linear regression analysis showed the emission rate of pollutants is negatively correlated with the cycle length and positively correlated with average delay and stopping; however, adding stopping rates as additional independent variables does not significantly improve the model and the correlation is only significant in case of CO and CO2 even though the relationship is statistically significant for all pollutants.


vehicular emission, Cycle lengths, optimization, MOVES, VISSIM, signal timing

Document Type




Degree Name

Civil Engineering

Level of Degree


Department Name

Civil Engineering

First Committee Member (Chair)

Haobing Liu

Second Committee Member

Nicholas Ferenchak

Third Committee Member

Claude Morelli