Citation

Yu L, Li Z, Bill AR, Noyce DA. Transp. Res. Rec. 2015; 2485: 16-25.

Copyright

(Copyright © 2015, Transportation Research Board, National Research Council, National Academy of Sciences USA, Publisher SAGE Publishing)

DOI

10.3141/2485-03

PMID

unavailable

Abstract

The absence of roadway lighting during the night results in a disproportionally large number of traffic crashes and fatalities. Past research on safety effects of roadway lighting involved mostly qualitative studies and focused on intersections and rural roads rather than freeways and interchanges. However, it is desirable to quantify the safety benefits of lighting for cost-benefit analysis and eventual decision making on lighting installations. This paper aims to develop nighttime crash prediction models for roadway lighting in the form of safety performance functions (SPFs). A pilot study on all freeways and expressways in the jurisdiction of Dane County, Wisconsin, was conducted. The crash data collection covered five years' nighttime crashes on all freeways, expressways, and interchanges in Dane County. Nighttime crashes were identified and light-pole data were collected along with traffic and geometry data. On the basis of the data, six SPFs were developed for freeway segments and interchange ramps under lighted and unlighted conditions.

RESULTS showed that the number of lanes was a significant factor for interchange ramps. Entry ramps had significantly higher numbers of crashes than did exit ramps under unlighted conditions. Similar results were observed for interchange freeway segments and noninterchange freeway segments. The international roughness index, used as a surrogate for pavement friction values, was significant for all freeway segments' SPFs. Severity distribution functions were also developed.

RESULTS showed that the largest safety benefit could be achieved by installing lighting at interchange segments, entry ramps, or exit ramps with annual average daily traffic greater than 10,000 vehicles per day.