Safety impacts of geometric design on freeway segments with closely spaced entrance and exit ramps

The objective of this study was to evaluate the effects of geometric design on crash risk on freeway segments with closely spaced entrance and exit ramps. Traffic flow, geometric design features and crash data from 80 segments on 14 freeways in the state of California, United States were applied. A multilevel logistic regression model with cross-level interactions was developed, where traffic variables were put on the case level, and their estimated coefficients were defined as a function of geometric design variables on the segment level. A basic logistic model and a multilevel logistic model without cross-level interactions were developed for comparison. The result shows that the one with cross-level interactions provides the best goodness of fit. The results indicate that six categories of geometric design variables are significantly associated with crash risk, i.e. lane configuration, basic number of lanes, ramp spacing, theoretical gore, inner shoulder width and speed limit. All but one (inner should width) geometric design variables have significant interaction terms with traffic flow variables. The effects of geometric design variables on crash risk are not fixed but vary with traffic conditions. The findings of this study can provide design guidance to improve road safety of freeway segments with closely spaced entrance and exit ramps.