Technical Reports

Analysis of the Safety Impact of Train Horn Bans at Rail-Highway Grade Crossings: An Update Using 1997-2001 Data

  • 13
  • Aug
  • 2003
REPORT NUMBER: Contract No. DTFR53-03-P-00005
SUBJECT: Highway-Rail Grade Crossings, USDOT National Highway-Rail Crossing Inventory
KEYWORDS: Train Horn Rule and Quiet Zones
ABSTRACT: This report updates Westat's prior report to FRA (Zador, 2002) in terms of both data and statistical methods. Safety impact estimates in the prior report were based on 5-year accident data for 1992-1996. In this report, safety impact estimates are based on recent grade-crossing accidents during the 5-year period between 1997 and 2001. As in the prior report, this report also derives safety impact estimates by comparing observed accident frequencies to expected accident frequencies. In the prior report, FRA's methods (U.S. DOT, FRA, 1987; U. S. DOT, FRA, 2000) were used, in a slightly modified form, to estimate the expected accident frequency at each grade crossing from the grade crossing's physical and operating characteristics. In Step 1 of this three step procedure, FRA's basic accident prediction formula (without adjustment for recent local accident history, U.S. DOT, FRA, 1987) was used to predict expected accidents. In Step 2, grade crossings were grouped on predicted accident risk into 10 groups with roughly comparable accident risks. In Step 3, general linear models (GLM, see McCullaghand Nelder, 1989) were used to estimate overall percent difference (across all risk groups but by warning device class) between grade crossings in the same accident risk group with and without the ban.
Pdf Download PDF Document [467.5 KB]