Accounting for nonrandom arrivals in estimate of delay at signalized intersections

The primary objective of this research was to determine the effect of nonrandom or platoon arrivals on the estimate of delay at signalized intersections. The delay model used in the 1994 Highway Capacity Manual (HCM) accounts for nonrandom arrivals through the variable m, which can be shown to be equal to 8 kI, where k describes the arrival and service distributions at the intersection and I describes the variation in arrivals due to the upstream intersection. The 1994 HCM delay model m-values are a function of the arrival type, where the arrival type describes the quality of progression at the intersection. Although an improvement to the fixed kI-value used in the 1985 delay model, the 1994 m values are based on empirical studies from limited field data and do not account for the decrease in random arrivals as the volume approaches capacity at the downstream intersection. This research provides an estimate of the variable kI for arterial conditions. An analytical equation was developed as a function of the degree of saturation, and a separate equation was developed for each signal controller type. The results from this research show that the proposed kI's provide delay estimates closer to the measured delay compared with the delay estimates using the kI-values in the 1994 HCM delay model.