Road geometry-related crash distribution, economic losses, and an alternative severity level analysis approach using combined parameters utilizing relative percentage share

Road traffic crashes interact with the road’s geometry formation. Analyzing traffic crash, accident, and crash cost distribution is not sufficient to define the severity level of the road unless the combined effects are explored. The objective of this study is analyzing road severity level based on an alternative combined parameter approach using traffic crash, accident, and crash cost relative percentage share. The study used five-year traffic crash data from Budapest. In the study area, on average, around 45 people die annually due to traffic crashes. Based on the existing and an alternative severity indexing approach, even though the road bend was a severe segment, greater than 60% of traffic crashes were registered on the straight road. But an alternative approach showed that the severity level of straight roads was higher than cross roads, and the severity value of cross roads was reduced. In addition to that, all road segments exist under a moderate severity condition with a severity score range (1 < SS ≤ 2). On average, Hungary loses around 0.2% of GDP annually due to traffic crashes registered in the Budapest city. Utilizing the relative percentage share of road traffic crashes, accidents, and crash costs and the combined effects had a significant impact in defining severity level analysis. This study identified that having a higher percentage share of road traffic crashes, accidents, and crash costs doesn’t designate the severity level and vice versa. Poor pavement quality and over speeding make the bend road more severe. As a result, it is better to investigate the quality of the road pavement, and reducing the speed below the design limit (< 50 km/h) will bring safer traffic movement. At the end, the study recommends that using the stated alternative severity indexing approach was potentially more comprehensive than the existing methods.