This paper proposes a simulation-based approach to conduct stress tests on transportation systems subject to extreme scenarios of rainfall leading to flooding. Stress tests represent situations where at least one part of the system, e.g., hazard intensity, performance of assets, is significantly worse than expected. To conduct them, the proposed approach features a set of interacting models that capture the behavior of the system under the effect of the conditions imposed by the stress tests. These include models that capture the occurrence of hazard events, performance of infrastructure assets and network, and the societal impacts. The proposed approach was used to conduct stress tests on a road network in Switzerland and three types of stress tests were conducted, labeled climate change, which investigates performance with increases in rainfall intensity in the future; travel demand, which investigates increases in demand for travel due to societal developments; and restoration capacity, which investigates decreases in post-hazard restoration capacity. The results provide significant insight into the vulnerabilities of the system under the considered stress tests. This information can be used to better plan measures to improve the resilience of the system.

14th International Conference on Applications of Statistics and Probability in Civil Engineering (ICASP14)