Sensitivity analysis of road tunnel resilience through data-driven stochastic simulation

Abstract

A probabilistic resilience model for tunnels exposed to disruptive events is vital to understand and estimate the functionality loss and its recovery time due to these events. Performing sensitivity analysis will help identify the critical parameters contributing to tunnel resilience. This paper aims to identify tunnel resilience’s sensitivity for parameters like traffic volume, fire suppression system, changes in maintenance, and operation parameters using a simulation model that estimates overall tunnel resilience for a given period. Overall universal compatibility of the simulation model is checked for twenty-two tunnels using information from U.S. National Tunnel Inventory (NTI), and resilience correlation is established. The results show that resilience loss due to fire and accidents are directly correlated with traffic volume. A significant reduction in the loss due to fire can be seen from installing a fire suppression system. Increasing the service life of equipment and frequency of inspection and repair contributes to an increase in tunnels’ resilience index. Resilience correlation study for the twenty-two tunnels showed that an average resilience index for these tunnels is 96.57%. Linear correlations between tunnel length and the traffic loss due to fire and operation can be made. Accidents and fire events are correlated with average traffic in the tunnel. Tunnel speed limit, age, number of lanes, and bores do not show a considerable effect on disruptive events. Overall, the study shows that the proposed simulation model can encompass various disruptive events to estimate the resilience of the tunnel.