Delay Models at Signalized Intersections Considering Short Right-Turn Lanes and Right Turn on Red

The most widely used procedures for estimating delay at signalized intersections are provided by the Highway Capacity Manual (HCM). Despite many revisions, the current HCM delay models lack consideration of short turn lanes and vehicles making a right turn on red (RTOR), both of which are commonly seen in real-life situations. Neglecting the delay caused by short-lane blockages and excluding RTOR traffic, the existing models can underestimate the delay for the signalized intersections.;To eliminate these shortcomings of the current HCM, comprehensive delay models were developed by considering the chance of short-lane blockage and right turns proceeding during red. The proposed analytical models were constructed via two different approaches: probabilistic theories and queue accumulation polygons (QAP). They were capable of handling delay estimation given different inputs. The effectiveness and validity of the proposed models were verified by the microscopic simulation model, SimTraffic, which was calibrated based on intersection field data.;The proposed models, developed with two different ideologies, showed their distinct advantages. By following the traditional HCM procedures, the probabilistic model only requires basic information as inputs, including traffic volume, saturation flow rate, signal timing plan, and short-lane length. It enhanced the HCM models by being able to handle delay estimation for short right-turn lane and RTOR. By incorporating the RTOR queuing and discharging information, the QAP model overcomes the shortcoming of HCM models where RTOR are directly removed before the capacity or delay analysis. It is better at modeling vehicular arrival and departure for both through and right-turn traffic under various signal control strategies and blockage scenarios.;Using the probabilistic model, it was found that the length of the short right-turn lane strongly influences the intersection delay. Delay decreases as the length of the right-turn lane increases. RTOR can reduce the over-saturation, thereby reducing the approach delay substantially, especially when the length of the lane is insufficient to accommodate queuing vehicles. Based on the field data studied in this research, both QAP model and probabilistic model yield better estimates for the approach delay than the HCM model. The results also confirmed that the HCM tends to underestimate the approach delay by ignoring the short-lane blockage and delay contributed by RTOR traffic.