To examine the relationship of ramp design to truck accident rates, this paper present an analysis of truck accidents in Washington State, plus a comparison to limited data from Colorado and California. The freeway truck accidents are grouped by ramp type, accident type, and by four conflict areas of each merge or diverge ramp. These groups are then compared on the basis of truck accidents per location and per truck-mile of travel. The truck accident frequencies and rates were found not significantly different by ramp type alone, but were significantly different by conflict area and accident type, both between and within ramp types. It was also found that high volume ramps had lower rates of truck accidents per truck-mile of travel. Thus, a ramp's safety risk is related to accident type and conflict area, but not directly to truck volumes, which affects assessments of high-risk locations. Specifically, a ramp with few accidents but a high proportion of rollovers in the merge area may have a deficiency, or a ramp with a low accident rate per truck trip due to high truck volumes may still be a high-risk site. The paper describes a straitforward use of the accident data analysed in this manner to identify accident-prone sites for further investigation. (A)
Abstract