| description abstract | Safety issues along urban expressways have gained increasing attention for their direct impact on traffic operation, especially after a substantial amount of reconstruction work that altered the cross section designs to create additional travel lanes. Despite the obvious benefit of higher capacity, the safety effects related to the variety of lane widths are uncertain. This paper examines the impact of lane widths on crash frequency for various types of crashes based on field data obtained from the expressway system in Shanghai. Four data sets were manually collected, including crash data, traffic flow data, lane width data, and geometric design data. Based on these data, safety performance functions (SPFs) were developed using negative binomial regression models. Crash modification factors (CMFs) were built upon the SPFs to quantify the impact of various lane widths on crash frequency for each crash type of interest. Specifically, we considered five cases in this paper: total crashes, two-vehicle crashes, multivehicle crashes, crashes under congested flow conditions, and crashes under noncongested flow conditions. In addition, lane width data are grouped into three categories in our models: undersized lanes (average lane width ≤3.25 m), standard-sized lanes (average lane width around 3.45 m), and oversized lanes (average lane width ≥3.75 m). The standard-sized lanes are always found to have the lowest crash frequency among all five cases; both undersized and oversized lanes are found to increase crash frequency, that is, total crash frequency for undersized and oversized lanes increased by 1.9 times and 1.34 times, respectively. | |
