| Roadway improvements are often made based on safety or operations. While each factor is important, consideration of one factor may impact another negatively. The tradeoffs between factors are often not well understood. The failure or inability to consider these impacts is of particular concern, as a general improvement in one area (ex. operations), may have adverse results on another.;This dissertation presents a methodology to link safety and operations in order to determine the tradeoffs of improvements. A zero-inflated Poisson model was selected to model safety and was specified within a hierarchical Bayesian framework. Bayesian p-values were utilized to establish model confidence and found that the model which was developed appeared to fit the data reasonably well.;Model results indicated that the approach speed limit, opposing approach speed limit, right turn lane length, maximum green time for the through movement, median width, land use and left turn movement all had an effect on the number of crashes per site per approach at different levels of severity. Protected/permitted phasing and through movement had an effect on crashes at all types of severities. Protected/permitted phasing had an effect on the overall number of crashes per site by approach when not considering severity.;Twenty three case studies were simulated to determine how safety or operational countermeasures impacted their counterpart. Impacts were assessed at the approach level and for the intersection as a whole. Benefit-Cost analysis was performed to further quantify the financial impacts of selected countermeasures.;Based on the simulations, the impact of safety changes on operations could be considered to be more negative than positive. Benefit-Cost analysis found that, for safety improvements, only four of the ten study sites produced a B/C ratio greater than 1.0. The impacts on safety of operational countermeasures were generally found to be positive. Eleven of the thirteen operational study sites produced a B/C ratio greater than 1.0. |
