Estimation of delay at signalized intersections for oversaturation and variable demand conditions

Abstract

Measurement of delay is not always feasible due to the cost and effort involved. Estimation of delay using analytical models is an effective alternative. The most popular and widely used model is the Highway Capacity Manual's delay model. This model is deficient in modeling delay for conditions with variable demand, for variable time periods, and for conditions with a queue at the start of the time period of analysis. This thesis validates the Fambro-Rouphail delay model which is proposed as an alternative to the Highway Capacity Manual's delay model. The methodology involved the simulation of specific conditions of operation for a signalized intersection using the TRAF-NETSIM program and comparison of the delays thus generated, to the estimates by the proposed model. The simulation was conducted for three consecutive time periods with the demand, and the period of analysis being varied for every case. Conditions with queues at the start of the period of analysis were also simulated. The maximum total delay obtained from the simulation was statistically compared to the estimates by the model using linear regression. The results from this study indicate a strong linear relationship between the simulation results and the model estimates. Hence, it is concluded that the model is an effective alternative to the Highway Capacity Manual's delay model.