Standard and Nonstandard Roadway Lighting Compared with Darkness at Rural Intersections

Abstract

This report evaluates the safety aspects of roadway lighting at rural and near-urban three-way and four-way junctions by comparing unlit intersections with those lit with two different types of lighting: ( a) standard intersection lighting provided by the Ministère des Transports du Québec, Canada, and ( b) nonstandard lighting provided by the local municipalities. A night–day accident rate ratio was used to estimate the accident rate reduction for three categories of severity: fatal and personal injury accidents, property damage only accidents, and all accidents. Sites were selected with two sampling modes. The objective mode selected sites according to the accident thresholds, and the arbitrary mode systematically selected all sites with standard lighting. The night–day accident rate ratio was measured for 376 sites by dividing the annual average number of accidents (6,546) with an annual average traffic flow (760 billion vehicles), calculated for both night and day. The accident rate reduction, expressed as a percentage, was tested for validity with the Student's t-test at the 5% p-level. The results were split into 49 categories with 20 variables to ensure that no significant variation existed in the accident rate reduction related to a specific roadway condition or environment. Rural lighting of an intersection significantly reduced the night accident rate by 29% for nonstandard lighting and by 39% for standard lighting, in comparison with darkness. When the two sampling modes were compared, standard lighting reduced the night accident rate of nonstandard lighting by 29%, significant at the 5% p-level, when only objective data in the sampling were used.