Impact of Intersection Angle on Highway Safety

Federal Highway Administration2021PDFTech ReportHarkey, David L.Lan, BoSrinivasan, RaghavanKumfer, WesleyCarter, DanielNujjetty, Anusha patelUnited States. Federal Highway Administration. Office of Safety Research and DevelopmentUniversity of North Carolina (System). Highway Safety Research CenterTan, CarolEigen, Ana MariaUnited States. Federal Highway Administration. Office of Safety Research and DevelopmentUnited StatesIntersection elementsIntersectionsTraffic safetyCrash modification factorsIntersection angleIntersection safetyCrash modification factorCrash modification functionSkew angleNegative binomial regressionStop controlledFHWA-HRT-20-067DTFH61-11-C-00050Tan, Carol; Eigen, Ana MariaThe preferred design for conventional intersections includes adjacent legs that intersect at 90 degrees. However, there are occasions when physical constraints result in intersection angles less than 90 degrees and thus produce skewed intersections. Skewed intersections may create potential safety and operational problems for both motorists and non-motorists. To date, the research on problems related to skewed intersections has been limited, which may explain the lack of consensus among the policies and guidance that now exists in practice. The objective of this study was to derive quantitative relationships between intersection angle and safety for which intersection crashes define safety. The relationships were used to determine appropriate crash modification functions (CMFunction) for reducing or eliminating the skew angle of an intersection, determine if there is a critical minimum angle at which safety is substantially diminished, and assess the need for revising current geometric design policies and practices.Data-mining and -regression techniques were used to identify the most important predictor variables from many independent variables. Data from minor road stop-controlled intersections in Minnesota and Ohio were used to estimate negative binomial regression models using the identified variables. These models were used to derive CMFunctions for rural and urban three- and four-leg intersections. The relationship between the crash modification factor and intersection angle (represented by the CMFunctions) differs from previous studies. Agencies should consider modifying the minimum critical angle for intersections in roadway design policies to be revised to reflect the results of this research.109