Archived project

Assessing the Crash Risk Implications of Roadside Hazards

Goal: This study estimated the number of roadside hazards of different types along rural roads and developed some basic crash models for rural roads

Date: 1 January 2002 - 1 January 2004

Updates
0 new
0
Recommendations
0 new
0
Followers
0 new
2
Reads
0 new
4

Project log

Shane Turner
added a research item
A significant number of New Zealand rural (50%) and urban (27%) accidents involve a roadside hazard. The majority of these accidents (85% plus) are single vehicle loss-of-control accidents. Accidents involving roadside hazards are often more severe than accidents that do not involve such hazards. Between 20% and 28% of accidents involving a water body, tree or over bank are fatal or serious.Roadside hazard data were collected for over 850 km of rural roads, including 414 km of State highways. A number of hazard removal/relocation/protection scenarios (mass action programmes), were investigated in this study. The analysis focused on major hazard types: large trees, heavy poles and unprotected bridge piers as determined from analysis of the LTSA’s Accident Analysis System (AIS). Costs were estimated for removing all trees, poles and protecting unprotected bridge piers within 4.0 m of the edge of the seal on all rural roads within TLAs where such hazards were observed in significant proportions. Accident prediction models were developed using generalised linear modelling techniques. Models were developed for all roadside hazard accident types combined and disaggregated by injury severity for each of the main roadside hazard types, (e.g. poles and trees). The modelling indicated that traffic volume is the most important variable and, with all other variables held constant, a linear relationship between accident numbers and traffic volume gave the best fit to the data. Road width was also an important factor, with wider roads being associated with more roadside hazard accidents, probably due to higher speeds. However, the relationship between speed and road width has not been investigated in this research, with further research required in this area.
Shane Turner
added a research item
This paper discusses the impact of roadside hazards on the occurrence and severity of rural single vehicle accidents and the important causal variables. The accident prediction models developed so far indicate that the location and type of roadside hazards, the consistency of the horizontal alignment and seal width are important predictor variables. Summary data on the number, location and types of roadside hazards found alongside rural roads in New Zealand is also presented. The roadside hazard predictions are based on a random sample of road sections surveyed throughout New Zealand.
Shane Turner
added a project goal
This study estimated the number of roadside hazards of different types along rural roads and developed some basic crash models for rural roads