Analysis and Modeling of Measured Delays at Isolated Signalized Intersections

Vehicular delay is an excellent tool for evaluating the operation of signalized intersections. This paper presents a methodology for measuring and analyzing stopped delay as well as acceleration and deceleration delay at a traffic signal. Field data were collected at 12 screen lines (or stations) defining a “system” where randomly selected vehicles are traced for about 355 m. Crossing times collected at the screen lines were utilized to estimate vehicle speed, acceleration, and delay of selected vehicles. Results indicated that deceleration and acceleration lengths of “stopped” vehicles varied over a wide range, as did the deceleration and acceleration rates. The average deceleration–acceleration delay was estimated at 11.8 s/veh, which is in good agreement with most reported values in the literature. The delay due to nonstopped vehicles comprises only about 7% of the total delay and can be neglected, while delay estimated from stopped vehicles is about 93% of the total intersection delay. Relationships between delay components were modeled, and the analysis showed a ratio of about 2.0 between the total delay and stopped delay, a ratio that is significantly higher than the 1.3 factor adopted in the 1994 Highway Capacity Manual. The approach delay was estimated at about 85% of the total delay, indicating a significant percentage of total delay takes place downstream of the intersection stop line.