Development of a Procedure for Prioritizing Intersections for Improvements Considering Safety and Operational Factors

Improvements are required at many intersections in Florida to eliminate safety or operational deficiencies. Because the annual budget is limited, it is essential for the Florida Department of Transportation (FDOT) to implement a prioritizing procedure to select intersections for improvement to achieve effectiveness under financial restrictions. The existing prioritizing procedure is based on safety factor only by ranking the intersections according to benefit-cost ratio for safety without considering operational effectiveness. This report presents results obtained from a research project performed to develop a new procedure considering both safety and operational factors, therefore to generate a more reasonable priority list for programming intersection improvement projects. In this procedure, safety performance is evaluated by utilizing benefit-cost ratios obtained from the existing prioritizing method. Criteria of operational performance are delay reduction due to improvements and existing delay. These two criteria are estimated based on algorithms in HCM 2000. To combine safety and operational factors, the Multi-Layer Prioritizing (MLP) method is implemented with the three criteria. This method assigns each criterion to a layer according to its relative importance: a higher layer owns a higher importance level. Intersections are prioritized in a first layer, and then clustered into several groups based on their relative similarity. Within each group, intersections are ranked in a second layer, and grouped into subgroups again. The final priority list is derived from the ranking of intersections within each subgroup in a third layer. Two kinds of priority sequences are adopted in this study. One indicates that safety effectiveness has precedence over operational effectiveness; meanwhile another one represents an opposite effectiveness priority. A case study was performed in this research project to demonstrate the new prioritizing procedure. Data from 34 intersections in District 7 in Florida were collected for operational analysis. These data included peak hour volume, traffic control, and geometric design. Two kinds of priority lists were produced as a result of implementation of the new procedure: safety preferred and operations preferred. A comparison between these priority lists and lists based on existing procedure was conducted finally to evaluate merits of the new procedure