PREDICTING ACCIDENT RATES OF RURAL ROADS

This paper examines the techniques used to predict the accident rates of sections of rural roads. The paper examines the qualities of a good road design and the means currently used in Australia to evaluate whether a safety barrier is required at a particular location. The paper introduces a cost-benefit analysis technique and then describes how this may be used in practice. Finally the paper compares the predicted collision rates using the techniques described in AASHTO (1977) and AASHTO (1989) with the measured collision rates on six rural roads around Brisbane. The findings were that the AASHTO 1977 method over-predicts the collision rates but this is likely to be due to differences in the vehicle speeds and the assumptions about what would be reported accidents. The AASHTO 1989 technique gave predicted values which were closer to the reported accident rates but the relationships were not as statistically significant. The conclusions reached were that the AASHTO 1977 technique or the AASHTO 1989 technique are acceptable and that they would be satisfactory to at least rank various sites. For the sites studies here, an average severity, estimated from the hazardousness of objects in the road side, can also be used to predict accident rates. (a) For the record of the covering entry of this conference, please see IRRD abstract no 868510.

Language

  • English

Media Info

  • Features: References;
  • Pagination: p. 85-97

Subject/Index Terms

Filing Info

  • Accession Number: 00729021
  • Record Type: Publication
  • Source Agency: ARRB Group Ltd.
  • ISBN: 0-86856-989-5
  • Files: ITRD, ATRI
  • Created Date: Dec 12 1996 12:00AM