Operational Analysis of Alternative Intersections

Alternative intersections and interchanges, such as the diverging diamond interchange (DDI), the restricted crossing u-turn (RCUT), and the displaced left-turn intersection (DLT), have the potential to both improve safety and reduce delay. However, partially due to lingering questions about analysis methods and service measures for these designs, their rate of implementation remains low. This research attempts to answer three key questions. Can alternative intersections and interchanges be incorporated into the existing level of service and service measure schema, or is a new service measure with an updated level of service model required? Is the behavior of drivers at alternative intersections fundamentally similar to those at conventional intersections, such that traffic microsimulation applications can accurately model the behaviors observed in the field? Finally, is the planning level tool made available through FHWA an accurate predictor of the relative performance of various alternatives, or is an updated tool necessary?

Discussion and case study analysis are used to explore the existing level of service and service measure schema. The existing control delay measure is recommended to be replaced with a proposed junction delay measure that incorporates geometric delay, with the existing level of service schema based on control type recommended to be replaced by a proposed schema using demand volume. A case study validation of micro- and macroscopic analysis methods is conducted, finding the two microscopic methods investigated to match field observed vehicle delays within 3 to 7 seconds for all designs tested, and macroscopic HCM method matching within 3 seconds for the DDI, 35 seconds for the RCUT, and 130 seconds for the DLT design. Taking the critical lane analysis method to be a valid measure of operations, the demand-volume limitations of each alternative design is explored using eighteen geometric configurations and approximately three thousand volume scenarios, with the DLT design predicted to accommodate the highest demand volumes before failure is reached. Finally, six geometries are examined using both the planning-level tool and the validated microsimulation tool, finding that the curve of the capacity-to-delay relationship varies for each alternative design, invalidating the use of critical lane analysis as a comparative tool.