Delay Variability at Signalized Intersections

Delays that individual vehicles may experience at a signalized intersection are usually subject to large variation due to randomness of traffic arrivals and interruption caused by traffic signal control. While such variation may have important implications for the planning, design and analysis of signal controls, currently there is no analytical model available to quantify it. This paper describes the development of an analytical model for predicting the variance of overall delay. The model is constructed on the basis of the delay evolution patterns under two extreme traffic conditions: highly undersaturated and highly oversaturated conditions. A discrete cycle-by-cycle simulation model is used to generate data for calibrating and validating the proposed model. The practical implications of the model are demonstrated through its use in analyzing factors influencing delay variability, in determining optimal cycle times with respect to delay variability, and in assessing level of service according to the percentiles of overall delay.